sports medicine, arthroscopy, rehabilitation, therapy & … · 2017-08-29 · biomed central...

BioMed Central

Sports Medicine, Arthroscopy, Rehabilitation, Therapy & Technology

ss
Open AcceResearchLower trunk kinematics and muscle activity during different types of tennis servesJohn W Chow*†1, Soo-An Park†2 and Mark D Tillman†3
Address: 1Center for Neuroscience and Neurological Recovery, Methodist Rehabilitation Center, Jackson, Mississippi, USA, 2Department of Orthopedic Surgery, Asan Medical Center, University of Ulsan, Seoul, South Korea and 3Department of Applied Physiology and Kinesiology, University of Florida, Gainesville, Florida, USA

Email: John W Chow* - [email protected]; Soo-An Park - [email protected]; Mark D Tillman - [email protected]

* Corresponding author †Equal contributors

AbstractBackground: To better understand the underlying mechanisms involved in trunk motion duringa tennis serve, this study aimed to examine the (1) relative motion of the middle and lower trunkand (2) lower trunk muscle activity during three different types of tennis serves - flat, topspin, andslice.

Methods: Tennis serves performed by 11 advanced (AV) and 8 advanced intermediate (AI) maletennis players were videorecorded with markers placed on the back of the subject used to estimatethe anatomical joint (AJ) angles between the middle and lower trunk for four trunk motions(extension, left lateral flexion, and left and right twisting). Surface electromyographic (EMG)techniques were used to monitor the left and right rectus abdominis (LRA and RRA), externaloblique (LEO and REO), internal oblique (LIO and RIO), and erector spinae (LES and RES). Themaximal AJ angles for different trunk motions during a serve and the average EMG levels fordifferent muscles during different phases (ascending and descending windup, acceleration, andfollow-through) of a tennis serve were evaluated.

Results: The repeated measures Skill × Serve Type × Trunk Motion ANOVA for maximal AJ angleindicated no significant main effects for serve type or skill level. However, the AV group hadsignificantly smaller extension (p = 0.018) and greater left lateral flexion (p = 0.038) angles than theAI group. The repeated measures Skill × Serve Type × Phase MANOVA revealed significant phasemain effects in all muscles (p < 0.001) and the average EMG of the AV group for LRA wassignificantly higher than that of the AI group (p = 0.008). All muscles showed their highest EMGvalues during the acceleration phase. LRA and LEO muscles also exhibited high activations duringthe descending windup phase, and RES muscle was very active during the follow-through phase.

Conclusion: Subjects in the AI group may be more susceptible to back injury than the AV groupbecause of the significantly greater trunk hyperextension, and relatively large lumbar spinal loadsare expected during the acceleration phase because of the hyperextension posture and profoundfront-back and bilateral co-activations in lower trunk muscles.

Published: 13 October 2009

Sports Medicine, Arthroscopy, Rehabilitation, Therapy & Technology 2009, 1:24 doi:10.1186/1758-2555-1-24

Received: 19 June 2009Accepted: 13 October 2009

This article is available from: http://www.smarttjournal.com/content/1/1/24

© 2009 Chow et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

of 14(page number not for citation purposes)

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Abstract&list_uids=19825184

http://www.smarttjournal.com/content/1/1/24

http://creativecommons.org/licenses/by/2.0

http://www.biomedcentral.com/

http://www.biomedcentral.com/info/about/charter/

Sports Medicine, Arthroscopy, Rehabilitation, Therapy & Technology 2009, 1:24 http://www.smarttjournal.com/content/1/1/24

IntroductionLow back injuries are common among competitive tennisplayers [1-7]. General agreement exists that mechanicalstress to the spine is related to the development of degen-erative disc disease in the lumbar region [8]. Among dif-ferent tennis strokes, the serve may place more stress onthe lumbar spine than the other strokes because repetitivetrunk hyperextension is generally thought to be the pre-disposing mechanism of spondylolysis [8-10]. Tennisplayers may be at an increased risk of lumbar disc pathol-ogy from rotational and hyperextension shearing effects[2]. The three types of serves that are widely used in tennisare the flat (minimum spin), topspin, and slice (sidespin)serves. In general, the flat serve is associated with a moreforceful action and produces the fastest ball speed amongthe three types of serves. The spin involved in the topspinand slice serves permit the server to hit with greater accu-racy. The racquet movement pattern and ball contact loca-tion relative to the body are different among these serves[11].

Because the spine is a complicated structure composed ofmany segments, joints, discs and various supporting mus-cles to protect the spinal cord and support the trunkmobility, it is very difficult to pinpoint the anatomic struc-tures that cause low back pain. Identifying the biome-chanical pathophysiologic factors associated with lumbarspinal loads may help to explain and prevent low backpathology. In national and world class tennis players whohad structural disorders in their lumbar spines, Saal [12]reported a 3-to-1 ratio of disc to postelement syndromes.In addition, for the young tennis players, posterior ele-ment pain spondylolysis with or without spondylolisthe-sis compose the injury subset most frequently. Alyas et al.[13] found pars injuries and facet joint arthroses, predom-inately in the lower lumbar spine, to be relatively com-mon in elite adolescent tennis players.

Motions of the trunk during occupational tasks have beenidentified as potential risk factors for developing low backdisorders (LBD) in manual workers [14]. High values ofcombined trunk velocities (e.g., simultaneous lateral flex-ion and twisting velocities) were found to occur moreoften in high LBD risk jobs than in low LBD risk jobs.Dynamic strength of the trunk and structural loading areconsidered the two major contributing factors to the rela-tionship between trunk dynamics and LBD. Structuralloading factors include biomechanical factors that con-tribute to loading on the spinal structures such as intra-abdominal pressure, muscle activity, and the imposedtrunk moment, and the actual loads on the structures ofthe spine.

Loading on the spinal structures can be very high during atennis serve. This is especially true when the racquet

moves behind the body and the vertebral column is later-ally flexed and hyperextended. Acceleration of the racquetbefore ball impact is accompanied by a rapid reversal ofthe rotation of the lumbar spine - from hyperextension toflexion and right twist to left twist for a right-hander. Thiscork-screwing motion transfers the force of its torque tothe spinal segments [15].

Body segmental and racquet kinematics of the tennis servehave been investigated extensively [16-18]. However,lumbar spine kinematics during the tennis serve have notbeen reported. Because low back injury is one of the mostprevalent musculoskeletal diseases in tennis, a betterunderstanding of the underlying mechanisms involved intrunk motion during a tennis serve is needed.

Information that identifies the muscles involved in strokeproduction is important for coaches and physical trainers[19]. Activity of the trunk muscles can be used to speculateon the stress on the lumbar intervertebral joints duringdynamic tasks [20-23] and it has shown that the force gen-eration and muscle recruitment activities associated withtwisting change significantly as a function of the torsoposture [24]. Prior EMG analyses of the tennis serve havefocused on the muscles in the hitting arm, shoulderregion, and lower extremity [25-31]. Very limited data onthe activity of the lower trunk muscles during the tennisserve are available. Anderson [32] reported that both theleft and right external obliques were very active (greaterthan 50% of the muscle's peak level of activity) during theforce production phase of the tennis serve. In a prelimi-nary study, Chow et al. [33] examined the muscle activa-tion patterns of eight lower trunk muscles during flat,topspin, and slice serves in five male highly skilled tennisplayers. They found no major differences in muscle activa-tion pattern across different serve types, and bilateral dif-ferences in muscle activation were more pronounced inrectus abdominis and external oblique than in internaloblique and lumbar erector spinae muscles. An apprecia-ble amount of abdominal/low back and bilateral co-acti-vation was observed during certain phases of the serve.

The purpose of this study was to evaluate the middle andlower trunk kinematics and selected lower trunk muscleactivity during three different types of tennis serves (flat,topspin, and slice) in skilled male players. Because playersstruck the ball at a much more backward and to the left(for a right-hander) location on their second as comparedto their first serve [11], players may need to arch backwardand laterally flex more when executing a topspin serve.Therefore, it was hypothesized that greater spinal range ofmotion and lower trunk muscle activity would be foundin the topspin serve when compared to the other servetypes.



MethodsSubjectsEleven advanced (AV) (United States Tennis AssociationNational Tennis Rating Program (NTRP) 5.5, age 25.3 ±4.1 years, height 180.3 ± 5.2 cm, mass 80 ± 8 kg) and eightadvanced intermediate (AI) (4.5 - 5.0, 23.4 ± 6.5 years,180.0 ± 9.5 cm, 78 ± 7 kg) male tennis players served asthe subjects. NTRP ratings range from 1.0 (beginner) to7.0 (world class professional). All subjects were right-handed and were in good physical condition and free ofinjury at the time of participation. The NTRP ratings wereself-reported ratings. All subjects signed informed consentdocuments before their participation.

Experimental SetupFour gen-locked video cameras (60 Hz) were stationedbehind and to the left of the baseline of an indoor tenniscourt (Figure 1). One of the cameras was used to capturethe whole body and racquet movements. The other threecameras captured the locations of reflective markerslocated on the back of the subject. A Peak event synchro-nization unit was used to synchronize the video and elec-tromyographic recordings (Peak PerformanceTechnologies, Inc., Inglewood, CO). For the purpose ofspatial reference, a Peak calibration frame (1.5 m × 1.4 m× 1.3 m object space, 16 control points) was videotaped atthe beginning of each data collection session. The framewas positioned at the baseline where the trunk of the sub-ject was located during testing.

Data CollectionEach collection session started with EMG trials followedby kinematics trials. Separate trials for EMG and kine-matic data were needed because EMG electrodes inter-fered with reflective marker placement on the lower back.

EMG TrialsAfter jogging on a treadmill for five minutes as a warm-up,surface electrodes were attached to selected muscles of thelower trunk including the left and right rectus abdominis(3 cm lateral to the umbilicus), external obliques (approx-imately 15 cm lateral to the umbilicus), internal oblique(below the external oblique electrodes and just superior tothe inguinal ligament), and lumbar erector spinae (3 cmlateral to L3 spinous process) [22]. The skin surfaceswhere the electrodes were located were cleansed with alco-hol and shaved when necessary. Electrodes were placedover the bellies of each muscle parallel to the muscle's lineof action with a center-to-center distance of 2.5 cm. Usinga MESPEC 4000 telemetry system (Mega Electronics Ltd.,Kuopio, Finland), the EMG signals were preamplifiedwith a gain of 500 and band pass filtered at 8-1500 Hz(CMRR > 130 dB) close to the electrodes and telemetric-ally transmitted to a central receiver (gain = 1, Butterworthfilter, 8-500 Hz band pass). The amplified EMG signals

were sampled at 900 Hz (12-bit analog-to-digital conver-sion) using the Peak Motus system.

To obtain maximum EMG levels, two maximal isometriccontractions were performed before the experimental tri-als - the bent-knee sit-up with the trunk inclined atapproximately 30° to the horizontal and the trunk hyper-extension performed in the prone position on a treatmenttable. In both maximal contractions, the feet were con-strained and the resistance was applied manually at theshoulders. One trial was performed for each maximal con-traction and each trial lasted for about 5 s.

Overhead view of the experimental setupFigure 1Overhead view of the experimental setup.



After the isometric trials, the EMG transmitter was securedto the left hip of the subject using elastic bands. The sub-ject was then asked to perform three different types ofserves - flat, topspin, and slice. In all trials, the subjectserved with efforts comparable to his first serves duringcompetition. They were asked to target their serves at thecorner near the center line (flat and topspin) and sideline(slice). Seven trials were performed for each serve type andthe serve type was presented in a random order. At the endof each trial, the subject was asked to rate his own per-formance based on the pace of the ball and landing loca-tion using a 5-point scale (5 = excellent, 0 = poor). Foreach trial, the EMG signals were collected for 5 s.

Kinematic TrialsAfter the EMG trials, electrodes were removed and eightreflective markers (each 1 cm in diameter) were placed onthe back of the subject. The marker locations were rightand left tips of the 11th rib, T9 and T12 spinous processes,right and left posterior superior iliac spines (PSISs), andL3 and L5 spinous processes (Figure 2). The location ofthe spine level was estimated using the techniques pro-posed by Tully and Stillman [34] and Schache et al. [35].These markers were used to estimate the 3-dimensional(3D) orientations of the spinal regions directly above andbelow the lumbar spinal segments. These two regionswere considered the middle and lower trunk for the pur-pose of this study. As a result, the change in relative

motion between the middle and lower trunk was treatedas motion in the lumbar spinal segments.

To establish the neutral orientation for the markersattached to each subject, the subject was asked to adopt aself-selected comfortable standing posture with armsfolded in front of the torso while marker locations wererecorded. The subject was then asked to perform seven tri-als for each type of serve and the order for the type of servewas assigned randomly. Again, the subjects served withefforts comparable to their first serves during competitionand rated their own performance at the end of each trial.

Data ReductionFor each subject, the two highest rated EMG and kinemat-ics trials for each type of serve were analyzed. For each trialbeing analyzed, four critical instants were identified fromthe video recordings of whole body and racquet move-ments: (a) beginning of the windup, the instant when theracquet passed in front of the legs, (b) end of the ascend-ing windup, the instant when the racquet reached thehighest position during the windup, (c) end of the win-dup, the instant when the racquet reached the lowest posi-tion behind the trunk, and (d) ball impact. For thepurpose of this study, the serve was divided into fourphases (Figure 3):

• Ascending windup [instants (a) to (b)]

• Descending windup [instants (b) to (c)]

• Acceleration [instants (c) to (d)]

• Follow-through [0.1 s duration after instant (d)]

These phases are of interest because of the distinct move-ment and functional characteristics of the body and rac-quet during these phases.

EMG DataThe raw EMG signals were filtered using a recursive digitalfilter (Matlab Elliptic filter, 10-450 Hz band pass) andfull-wave rectified. The maximum isometric trial datawere smoothed using a moving average of 2 s and the larg-est average EMG value recorded for each muscle was con-sidered the maximum EMG level. The experimental trialdata were smoothed using a moving average of 50 msbefore normalizing to the respective maximum EMG lev-els. An average normalized EMG value was computed foreach muscle in each phase for each trial analyzed.

Kinematic DataFor each standing or serving trial analyzed, coordinatedata were extracted from the video pictures (automatictracking) using a video-based motion analysis system

Reflective marker locations and coordinate systems for the middle and lower trunkFigure 2Reflective marker locations and coordinate systems for the middle and lower trunk.



(Peak Motus Motion Measurement System). The three-dimensional (3D) coordinates the eight reflective markerslocated on the back of the trunk were transformed fromthe Peak reference frame to local reference frames embed-ded in the middle and lower trunk (Figure 2). Consideringthe middle and lower trunk as adjacent segments of ajoint, the anatomical joint (AJ) angles between the twosegments is the relative orientation of the two segment-embedded local reference frames (see Additional file 1,[36]). The three components of the AJ angles represent therotations about the medio-lateral axis (flexion/extensionangle), antero-posterior axis (lateral flexion angle), andlongitudinal axis (twisting angle). For each subject, the AJangles obtained during serving trials were expressed as theangular deviation from the AJ angles recorded at standingposture (i.e., the AJ angles at standing posture are all 0°).The dependent variables for trunk motion were the maxi-mal extension, left lateral flexion, and left and right twist-ing angles during a serve. For each subject, averagenormalized EMG values and maximum AJ angles over twotrials of the same serve type were used for subsequentanalysis.

Data AnalysisFor each serve type, phase and skill level combination,mean and standard deviation were computed for each var-iable of interest. For each muscle, the EMG parameterswere compared using a 2 × 3 × 4 (Skill × Serve type ×Phase) multivariate analysis of variance (MANOVA) withrepeated measures on the last two factors. Separate univar-iate tests were performed for follow up testing whenappropriate, and Bonferroni's procedure was used toadjust the overall type I error rate. To determine if signifi-cant variations existed among skill groups, serve types andtrunk motions in the maximal AJ angle, a 2 × 3 × 4 (Skill× Serve type × Trunk motion) ANOVA with repeatedmeasures on the last two factors was performed. An a-pri-ori alpha level of 0.05 and an a-priori beta level of 0.20were used in this study.

ResultsMuscle ActivationAs expected, a significant main effect for the phase wasfound in the average EMG level in each of the musclesmonitored (p < 0.001). In addition, no significant maineffect for the serve type or inter-factor interaction was

Critical instants of a serve -- (a) beginning of the windup, (b) end of the ascending windup, (c) end of the windup, and (d) ball impactFigure 3Critical instants of a serve -- (a) beginning of the windup, (b) end of the ascending windup, (c) end of the windup, and (d) ball impact. Adapted from Chow et al. [33].



found in any of the muscles (p > 0.695). In general, thelower trunk muscles become active toward the end of theascending windup phase (Figure 4). For most musclestested, the largest average EMG levels were observed ineither the descending windup or acceleration phases.When comparing overall muscle activation during a ten-nis serve between the two skill groups, the subjects in theAV group generally exhibited greater muscle activationthan the subjects in the AI group (Figure 5).

Rectus AbdominisA significant main effect for the skill was found in the LRAactivity (p = .008). The AV group showed significantlyhigher LRA activation than the AI group. Regardless of theserve type, the activation patterns of the LRA and RRA arequite similar - activations in the descending wind-up andacceleration phases are greater than the ascending windupand follow-through phases (Table 1).

External ObliqueDifferent patterns of activation were observed in the twoEO muscles (Table 1). In general, the subjects in the AVgroup exhibited greater LEO activity than the subjects inthe AI group and the difference was a statistical trend (p =0.055).

Internal ObliqueAmong the muscles examined, the IO muscles have thelargest overall EMG levels (Figure 4). Different patterns ofactivation were observed in the two IO muscles (Table 1).The LIO was generally more active than the RIO through-out a serve except in the follow-through phase. Very highactivation levels (average EMG levels greater than 100%max) were found in the descending windup and accelera-tion phases in the LIO.

Average EMG levels of different muscles during different phasesFigure 4Average EMG levels of different muscles during different phases.



Erector SpinaeModerate activity was observed in all phases except theascending windup for both ES muscles and the descend-ing windup phase for the RES (Table 1). Different patternsof activation were observed in the two ES muscles (Figure4). Instead of a relatively constant average EMG level afterthe ascending wind-up phase observed in the LES, theactivity of RES increased steadily from ascending wind-upto follow-through phase. Although not statistically signif-icant (p = 0.069), the AV group showed greater RES acti-vation than the AI group (Table 1).

Maximum AJ AnglesThe repeated measures ANOVA performed on maximumAJ angles revealed a significant main effect for trunk

motion (p < 0.001). This simply means that the middle-lower trunk ROM is not the same in different principleplanes (Table 2). In addition, a significant interaction wasfound between trunk motion and skill factors (p < 0.001)(Figure 6). The AV group exhibited greater maximum AJangles in all trunk motions measured except the exten-sion. No significant main effect for serve type wasdetected.

The ANOVA also revealed significant differences betweenthe two skill levels for the extension and left lateral flexion(Table 2). For the extension motion, the AV group exhib-ited a significantly smaller maximum AJ angle than the AIgroup (p = 0.018) while the opposite was observed in theleft lateral flexion (p = 0.038).

Average EMG levels of different muscles for the two skill groupsFigure 5Average EMG levels of different muscles for the two skill groups.




Table 1: Average (SD) EMG (%MAX) during different phases for different serve types and muscles for subjects of different skill levels.

Flat Topspin Slice

Muscle Skill AWU DWU ACC FT AWU DWU ACC FT AWU DWU ACC FT

Left RA* AV 25(45)

62(34)

92(89)

28(28)

28(44)

59(40)

79(71)

25(24)

24(41)

70(59)

87(90)

26(16)

AI 12(12)

50(39)

47(22)

21(21)

10(7)

55(48)

64(35)

36(44)

15(13)

43(33)

41(28)

11(6)

Right RA AV 20(38)

37(43)

56(30)

20(11)

20(35)

24(16)

47(34)

15(7)

21(40)

33(28)

53(34)

26(23)

AI 8(4)

40(35)

59(34)

23(14)

10(6)

36(32)

65(50)

23(14)

18(26)

35(26)

60(43)

26(19)

Left EO AV 44(50)

121(86)

110(80)

5(40)

53(61)

100(68)

109(78)

52(35)

42(49)

115(100)

101(57)

46(24)

AI 30(30)

104(95)

67(56)

43(35)

25(23)

86(60)

89(65)

69(95)

32(29)

96(84)

67(50)

33(29)

Right EO AV 25(26)

54(32)

93(70)

47(36)

26(31)

55(44)

79(61)

37(27)

20(21)

59(36)

100(80)

42(23)

AI 18(10)

58(16)

88(60)

38(21)

17(10)

51(14)

93(61)

48(35)

18(7)

48(14)

72(53)

39(34)

Left IO AV 51(41)

183(144)

118(61)

59(38)

59(61)

150(94)

111(61)

52(36)

40(27)

168(127)

136(91)

73(68)

AI 42(43)

151(113)

91(49)

55(26)

31(24)

143(97)

138(84)

109(123)

48(35)

146(119)

108(64)

43(9)

Right IO AV 30(28)

69(38)

116(76)

88(92)

29(29)

81(60)

100(63)

66(60)

24(21)

82(58)

105(63)

76(76)

AI 25(25)

97(74)

100(63)

72(50)

26(27)

96(87)

93(64)

81(51)

28(25)

102(101)

89(55)

63(54)

Left ES AV 21(20)

51(28)

42(27)

41(19)

19(20)

47(29)

35(20)

37(17)

17(18)

48(28)

44(25)

42(23)

AI 14(14)

38(24)

45(24)

40(31)

12(11)

31(23)

44(20)

52(63)

13(11)

38(25)

48(22)

33(19)

Right ES AV 17(21)

34(34)

46(31)

72(65)

17(22)

38(37)

45(34)

67(66)

15(17)

31(39)

56(55)

99(142)

AI 9(4)

12(5)

46(28)

59(21)

11(6)

13(8)

58(30)

59(22)

13(9)

17(10)

46(16)

63(25)

Abbreviations: AWU - ascending windup, DWU - descending windup, ACC - acceleration, FT - follow-through, RA - rectus abdominis, EO - external oblique, IO - internal oblique, ES - erector spinae, AV - advanced, AI - advanced-intermediate.* Significant difference between AV and AI (p < 0.05).


DiscussionWhen executing a tennis serve, vigorous movement of thetrunk helps to generate as much angular momentum aspossible and transfer it to the racquet [37]. Dynamic sta-bility of the spine is essential to prevent low back dysfunc-tion [38,39] and is associated with sufficient strength andendurance of the trunk stabilizing muscles and appropri-ate activation sequencing of the trunk muscles [40]. Spi-nal stability is also increased with either an increasedcoactivation of antagonistic counteracting trunk musclesor an increased intra-abdominal pressure during the staticcondition [38,41]. The primary focus of this study was toexamine the role of lower trunk muscles in providingdynamic stability of the lumbar spine during a tennisserve and to speculate on the lumbar spine loads during atennis serve using lower trunk muscle activation and kin-ematics data.

Muscle ActivationThe hypothesis that greater activation levels in lower trunkmuscles would be found in topspin serves when com-pared to the other serve types was not supported by ourresults. In general, the activation patterns of differentmuscles during a tennis serve are comparable to thosereported by Chow et al. [33]. However, Chow et al. [33]used players with skill levels similar to the subjects in theAV group of the present study. For the purpose of this dis-cussion, an average EMG level of less than 10% max isconsidered low. Muscle EMG values of approximately50% and 100% max are considered moderate and high,respectively.

Rectus AbdominisThe RA muscles are active during trunk flexion motionslike curl-up, sit-up or leg raising exercises [42]. Due to itsvertical (longitudinal) alignment, the RA has minimal

Interactions between skill level and trunk motion in maximum anatomical joint angleFigure 6Interactions between skill level and trunk motion in maximum anatomical joint angle.



contribution to the production of torque in the transverseplane [43]. High activation level of anterior supportingmuscles of the lower trunk including RA may lead to unfa-vorable forces on the spine [44]. It has been suggested thatan average normalized EMG value to MVIC (maximal vol-untary isometric contraction) below 30% is not consid-ered very stressful to the spine structures [45]. During thedescending windup and acceleration phases of a tennisserve, the EMG level of both RA muscles increased to 40%or higher (Table 1). The AV group showed a higher levelof activation during the acceleration phase than the AIgroup. With the center of gravity of the upper bodylocated behind the lumbar spine, the RA activity (co-con-traction) in a hyperextended posture during the descend-ing windup and acceleration phases can drasticallyincrease the loads on the lumbar spine and lead to harm-ful stress to the lumbar spine structures.

External ObliqueThe EO muscle is one of the anterior supporting musclesof the lower trunk that is also active during trunk flexionexercises [42]. However, it has been well recognized thatthe contralateral EO muscle is one of the main movers foraxial rotation of the trunk [46,47]. An appreciable antag-onistic activity of the ipsilateral EO during axial rotationwas also reported in the literature [48]. In the presentstudy, the co-contraction of bilateral EO muscles helps tostabilize the lumbar spine during the tennis serve. Thistype of co-contraction helps to increase the compressiveload and lead to the torsional stiffness of the lumbar spinesegments [49,50]. High activation of LRA and LEO for theright-handers in the present study indicated that these

muscles acted as prime movers of twisting to the rightoccurred during the descending windup and early acceler-ation phases.

Internal ObliqueIpsilateral IO is the agonist to the contralateral EO for theaxial rotation of the trunk [46]. Similar to the EO muscles,an appreciable co-contraction of contralateral IO is com-mon during trunk twists. Interestingly, bilateral differencein muscle activation is more pronounced in IO than in theEO [51]. One explanation is that the function of EO ismore complicated than just acting as a prime axial rotatorof the lower trunk [48,52]. Like the EO muscles, bilateralco-contraction of IO observed during a tennis serve helpsto provide a stabilizing force to the lumbar spine.

Erector SpinaeThe lumbar ES lies lateral to the multifidus muscle andforms the prominent dorsolateral contour of the backmuscles in the lumbar region. The lumbar ES consists oftwo muscles - the longisimus dorsi and iliocostalis. Thelines of action of these two muscles are mostly vertical (orlongitudinal) and a bilateral ES contraction can act as aposterior sagittal extensor. However, when contractingunilaterally, these muscles can act as lateral flexors of thelumbar vertebrae [53]. During axial rotation, the ipsilat-eral ES is more active than the contralateral ES [54]. It alsohas been suggested that, during axial rotation, back mus-cles maintain the spinal posture and stabilize the lumbarspine [22,43].

In the present study, we found different patterns of muscleactivation during a tennis serve for the two ES muscles.The LES is quite active throughout a tennis serve exceptduring the ascending windup phase while sequentiallyincreased activity from ascending windup to follow-through phases was observed in the RES (Table 1). It isobvious that, for the right-handed subjects in this study,the LES assisted the lateral flexion to the left after theascending windup phase. To stabilize the trunk during anunbalanced posture in the follow-through phase, the RESbecomes highly active during the follow-through phase.The bilateral ES co-contraction is more pronounced in theAV group. This is probably related to the greater left lateralflexion found in the AV subjects (Figure 6). In addition tobilateral co-contractions, front/back co-contractions existthroughout a tennis serve and are especially high towardto end of a serve. This may suggest that the lumbar spineis subject to large compression loads during the follow-through phase.

Lower Trunk MotionIn-vivo techniques have been employed to measure spinekinematics during various physical activities [55-57].However, it is not feasible to use these techniques in the

Table 2: Mean (SD) maximum anatomical joint angles in degrees.

Serve type Flat Top-spin Slice

Motion* AV AI AV AI AV AI

Extension# 19.3(11.9)

27.5(10.9)

19.3(10.6)

31.9(17.2)

20.0(10.1)

26.9(7.6)

Left lateral Flexion# 16.0(4.1)

12.3(5.2)

15.5(5.4)

10.9(4.8)

15.4(5.3)

12.4(8.4)

Left Twisting 7.9(4.3)

6.8(3.8)

6.8(4.1)

5.4(0.96

5.5(3.8)

4.3(3.1)

Right Twisting 5.6(4.6)

4.2(1.8)

6.1(6.9)

4.1(0.6)

11.2(12.7)

4.1(3.1)

* Significant difference among different motions (p < 0.01).# Significant difference between AV (advanced) and AI (advanced intermediate) groups (p < 0.05).



present study because of the large ROM associated with atennis serve. Alternatively, we used markers placed on thelower back to estimate lumbar spinal motion during a ten-nis serve. The major limitation of our procedures is theskin movement relative to the spine during lower trunkmotions. Despite this limitation, the marker locationsallow for reasonable estimation of relative motionbetween the middle and lower trunk during a tennis serve.

ExtensionDuring the extension motion, the vertebral bodiesundergo posterior sagittal rotation and a small posteriortranslation. A downward movement of the inferior articu-lar processes and the spinous process is also involvedwhich limited by bony impaction between spinous proc-esses [58]. This type of impaction is accentuated when thejoint is subjected to the action of the back muscles [59].The maximal extension AJ angles obtained in the presentstudy fall within the extension ROM values reported in theliterature [56,60]. It should be emphasized that it may notbe adequate to compare the AJ angles in this study withthe ROM values reported by other investigators because ofthe differences in measuring techniques.

The reason why the AI group had significantly greatermaximum extension AJ angles than the AV group is diffi-cult to understand. One explanation is that, instead ofrelying on lumbar hyperextension like the AI subjects did,the subjects in the AV group relied more on the hyperex-tension of the upper trunk (i.e., thoracic spine) to achievethe overall trunk hyperextension needed for an executionof a tennis serve.

Left Lateral FlexionThe lateral flexion of the lumbar spine involves a complexand variable combination of lateral bending and rotatorymovements of the inter-body joints and diverse move-ments of the zygapophysial (facet) joints. When com-pared to lateral bending ROM values for lumbar spinalmotion segments measured by X-rays [60,61], the maxi-mal left lateral flexion AJ angles are close to the summedvalue of each motion segment in the lumbar spine. Thesignificantly greater maximal left lateral flexion AJ angleexhibited by the AV group indicates that highly-skilledright-handed players can reach for a greater height duringa tennis serve because of the greater left lateral flexion. Thesignificantly greater lateral flexion AJ angle corresponds tothe significantly greater LRA activity found in the AVgroup. This implies that highly-skilled players are sub-jected to greater asymmetric loads on their lumbar spinesdue to the greater lateral flexion.

Axial RotationAxial rotation of the lumbar spine involves twisting of theintervertebral discs and impaction of zygapophysial

joints. During axial rotation of an intervertebral joint, allthe fibers of the annulus fibrosus which are inclined to thedirection of rotation will be strained. The other half willbe relaxed. Based on the observation that an elongation ofcollagen beyond about 4% of its resting length can lead toinjury of the fiber, it has been estimated that the maxi-mum range of rotation of an intervertebral disc withoutinjury is about 3° [62]. The twisting ROM for each lumbarspinal motion segment ranges from 0 - 2° [61]. The max-imal twisting AJ angles found in the present study areslightly greater than the ROM values reported in the liter-ature. The maximal twisting AJ angles are consideredsmall compared to the amount of shoulder movementduring a tennis serve. This clearly indicates that the axialrotation of the trunk during a tennis serve is mostly fromthe twisting of the upper trunk.

Co-contractionsBoth bilateral and abdominal/back co-contractionsamong lower trunk muscles are unavoidable during trunkmovements because these muscles function as units tomaintain the balance between mobility and stability ofthe spinal column. As a result, the lumbar spine is sub-jected to a large amount of compressive and torsionalstress during athletic movements due to the co-contrac-tion. Although the importance of torsional stress in theetiology of disc degeneration and prolapse is inconclusive[63,64], the link between high compressive load and lowback injury and pain is well documented [65,66]. The acti-vation patterns of the lower trunk muscles clearly demon-strate a high degree of co-contraction during a tennisserve, especially in the descending windup and accelera-tion phases. In addition to the compressive load, thehyperextension and lateral flexion of the trunk during var-ious phases of a tennis serve may cause shear loads on thelumbar spine. Consequently, stresses upon the variousanatomical structures may result in spinal injury and backpain.

Practical ImplicationsThe risk of spinal injury can be high for tennis players ofdifferent skill levels. It is well known that physical activityincreases the amount of bone mineral in the skeleton[67,68]. Granhed et al. [69] found that intensive weightlifting would increase the bone mineral content in thelumbar vertebrae to an extent that the spine can tolerateextraordinary loads. Unlike most competitive tennis play-ers, recreational players usually do not spend much timeon training or supervised strength and conditioning pro-grams. As a result, their vertebrae are not as strong as thetrained individuals and are more susceptible to injurywhen subjected to large lumbar spinal loads. Althoughhighly-skilled competitive players are likely to havestronger vertebrae when compared to recreational players,they are also susceptible to spinal injury due to other fac-



tors. Because competitive players complete a largenumber of serves in practices and competitions, the accu-mulative stress on the lumbar spine can be detrimental.

One of the risk factors that has been overlooked by sportsmedicine practitioners is the possible link between the"time of occurrence" and back injury. Adams et al. [70]measured the range of lumbar flexion of human subjectsin the early morning and in the afternoon and the bend-ing properties of cadaveric lumbar segments before andafter creep loadings that simulate a day's activity. Theyconcluded that lumbar discs and ligaments are at greaterrisk of injury in the early morning compared with later inthe day. Although the focus of their study was on lumbarflexion, it does have implications to lumbar spinalmotion in general. It seems reasonable to advise patientswith history of back disorders to avoid activity that willput the lumbar spine in extreme range of motion such asthe tennis serve in the early morning. To gain more insightinto this issue, inclusion of the "time of occurrence" infuture epidemiological studies of acute low back injury isrecommended.

The heavy involvement of lower trunk muscles in the ten-nis serve reinforces the importance of abdominal andlower back exercises in the strength and rehabilitationprograms designed for tennis players. Because most lowertrunk muscles undergo eccentric contractions duringselected phases of the serve, it is recommended that eccen-tric training is included in the conditioning programs. Thestrengthening of the lower trunk muscles not only willenhance performance, but the tennis players will also ben-efit in preventing low back injury and pain.

Recommendations for future studies1. Future studies should examine if there are differences inactivation patterns of the lower trunk muscles during thetennis serve among players of different skill levels includ-ing beginners.

2. In addition to lower trunk muscles, other back musclessuch as multifidus and thoracic erector spinae muscles canbe examined.

3. The use of EMG techniques to identify the muscle acti-vation characteristics that are associated with low backpain [71] can be explored in tennis players.

4. Several epidemiological studies on low back disordersamong young competitive tennis players have been con-ducted. However, such data are not available for non-competitive players of different ages. Future effort shouldexamine the incidence of low back injuries in recreationalplayers.

AbbreviationsMuscles

RA: rectus abdominis (LRA: left RA; RRA: right RA); EO:external oblique (LEO: left EO; REO: right EO); IO: inter-nal oblique (LIO: left IO; RIO: right IO); ES: erector spinae(LES: left ES; RES: right ES).

Others

AJ angle: anatomical joint angle; ANOVA: analysis of var-iance; MANOVA: multi-variate analysis of variance; EMG:electromyography or "electromyographic"; ROM: range ofmotion.

Competing interestsThe authors declare that they have no competing interests.

Authors' contributionsJWC contributed to conception and design, acquisition ofdata, analysis and interpretation of data, and prepared themanuscript with the assistance of the other authors. BothSAP and MDT contributed to conception and design,acquisition of data, and revised the manuscript criticallyfor important intellectual content.

Additional material

AcknowledgementsThis research project was funded in part by the United States Tennis Asso-ciation (USTA). The authors wish to thank Kim Fournier, Guy Grover, Greg Gutierrez, Chris Hasler, Ryan Mizell, Dana Otzel, Joceyln Plesa, and Dileep Ravi for their assistance in different aspects of this project.

References1. Bastholt P: Professional tennis (ATP tour) and number of

medical treatments in relation to type of surface. Med Sci Ten-nis 2000, 5(2): [http://www.stms.nl/index.php?option=com_content&task=view&id=881&Itemid=263].

2. Hainline B: Low back injury. Clin Sports Med 1995, 14:241-265.3. Lundin O, Hellstrom M, Nilsson I, Sward L: Back pain and radio-

logical changes in the thoraco-lumbar spine of athletes. Along-term follow-up. Scand J Med Sci Sports 2001, 11:103-109.

4. Marks MR, Haas SS, Wiesel SW: Low back pain in the competi-tive tennis player. Clin Sports Med 1988, 7:277-287.

5. Sward L, Hellstrom M, Jacobsson B, Peterson L: Back pain and radi-ologic changes in the thoraco-lumbar spine of athletes. Spine1990, 15:124-129.

6. Vad VB, Gebeh A, Dines D, Altchek D, Norris B: Hip and shoulderinternal rotation range of motion deficits in professional ten-nis players. J Sci Med Sport 2003, 6:71-75.

7. Winge S, Jorgensen U, Lassen NA: Epidemiology of injuries inDanish championship tennis. Int J Sports Med 1989, 10:368-371.

Additional file 1Appendix. Computation of Anatomical Joint Angles.Click here for file[http://www.biomedcentral.com/content/supplementary/1758-2555-1-24-S1.PDF]


http://www.biomedcentral.com/content/supplementary/1758-2555-1-24-S1.PDF

http://www.stms.nl/index.php?option=com_content&task=view&id=881&Itemid=263

http://www.stms.nl/index.php?option=com_content&task=view&id=881&Itemid=263















8. Kelsey JL, White AA III: Epidemiology and impact of low-backpain. Spine 1980, 5:133-142.

9. Congeni J, McCulloch J, Swanson K: Lumbar spondylolysis. Astudy of natural progression in athletes. Am J Sports Med 1997,25:248-253.

10. Cyron BM, Hutton WC: Variations in the amount and distribu-tion of cortical bone across the partes interarticulares of L5.A predisposing factor in spondylolysis? Spine 1979, 4:163-167.

11. Chow JW, Carlton LG, Lim YT, Chae WS, Shim JH, Kuenster AF,Kokubun K: Comparing the pre- and post-impact ball and rac-quet kinematics of elite tennis players' first and secondserves: a preliminary study. J Sports Sci 2003, 21:529-537.

12. Saal JA: Tennis. In The Spine in Sports Edited by: Watkins RG. St.Louis, Missouri: Mosby; 1996:499-504.

13. Alyas F, Turner M, Connell D: MRI findings in the lumbar spinesof asymptomatic, adolescent, elite tennis players. Br J SportsMed 2007, 41:836-841.

14. Davis KG, Marras WS: The effects of motion on trunk biome-chanics. Clin Biomech 2000, 15:703-717.

15. Elliott BC: Biomechanics of the serve in tennis. A biomedicalperspective. Sports Med 1988, 6:285-294.

16. Elliott B, Marsh T, Blanksby B: A three-dimensional cinemato-graphic analysis of the tennis serve. Int J Sport Biomech 1986,2:260-271.

17. Elliott BC: Spin and the power serve in tennis. J Hum MovemStud 1983, 9:97-104.

18. Van Gheluwe B, Hebbelinck M: The kinematics of the servicemovement in tennis: A three-dimensional cinematographi-cal approach. In Biomechanics IX Edited by: Winter D, Norman R,Wells R, Hayes KC, Patla AE. Champaign, IL: Human Kinetics Publish-ers; 1983:521-526.

19. Reid M, Chow JW, Crespo M: Muscle activity: An indicator fortraining. In Biomechanics of Advanced Tennis Coaching Edited by: Elli-ott B, Reid M, Crespo M. London: International Tennis Federation;2003:109-137.

20. Cook TM, Mann S, Lovested GE: Dynamic comparison of thetwo-hand stoop and assisted one-hand lift methods. J SafetyRes 1990, 21:53-59.

21. Delitto RS, Rose SJ: An electromyographic analysis of two tech-niques for squat lifting and lowering. Phys Ther 1992,72:438-448.

22. McGill SM: Electromyographic activity of the abdominal andlow back musculature during the generation of isometricand dynamic axial trunk torque: implications for lumbarmechanics. J Orthop Res 1991, 9:91-103.

23. Seroussi RE, Pope MH: The relationship between trunk muscleelectromyography and lifting moments in the sagittal andfrontal planes. J Biomech 1987, 20:135-146.

24. Marras WS, Davis KG, Granata KP: Trunk muscle activities dur-ing asymmetric twisting motions. J Electromyogr Kinesiol 1998,8:247-256.

25. Buckley J, Kerwin D: The role of the biceps and triceps brachiiduring tennis serving. Ergonomics 1988, 31:1621-1629.

26. Girard O, Micallef JP, Millet GP: Lower-limb activity during thepower serve in tennis: effects of performance level. Med SciSports Exerc 2005, 37:1021-1029.

27. Kibler WB, Chandler TJ, Shapiro R, Conuel M: Muscle activation incoupled scapulohumeral motions in the high performancetennis serve. Br J Sports Med 2007, 41:745-749.

28. Morris M, Healy B, Pink M, Perry J, Jobe F: Electromyographicanalysis of elbow function in tennis players. Am J Sports Med1989, 17:241-247.

29. Ryu RK, McCormick J, Jobe FW, Moynes DR, Antonelli DJ: An elec-tromyographic analysis of shoulder function in tennis play-ers. Am J Sports Med 1988, 16:481-485.

30. Seeley MK, Uhl TL, McCrory J, McGinn P, Kibler WB, Shapiro R: Acomparison of muscle activations during traditional andabbreviated tennis serves. Sports Biomech 2008, 7:248-259.

31. Van Gheluwe B, Hebbelinck M: Muscle actions and ground reac-tion forces in tennis. Int J Sport Biomech 1986, 2:88-99.

32. Anderson MB: Comparison of muscle patterning in the over-arm throw and tennis serve. Res Q 1979, 50:541-553.

33. Chow JW, Shim JH, Lim YT: Lower trunk muscle activity duringthe tennis serve. J Sci Med Sport 2003, 6:512-518.

34. Tully EA, Stillman BC: Computer-aided video analysis of verte-brofemoral motion during toe touching in healthy subjects.Arch Phys Med Rehabil 1997, 78:759-766.

35. Schache AG, Blanch P, Rath D, Wrigley T, Bennell K: Differencesbetween the sexes in the three-dimensional angular rota-tions of the lumbo-pelvic-hip complex during treadmill run-ning. J Sports Sci 2003, 21:105-118.

36. Vaughan CL, Davis BL, O'Connor JC: Dynamics of Human Gait.Cape Town: Kiboho Publishers; 1999.

37. Bahamonde RE: Changes in angular momentum during thetennis serve. J Sports Sci 2000, 18:579-592.

38. Cholewicki J, VanVliet JJ: Relative contribution of trunk musclesto the stability of the lumbar spine during isometric exer-tions. Clin Biomech 2002, 17:99-105.

39. Vezina MJ, Hubley-Kozey CL: Muscle activation in therapeuticexercises to improve trunk stability. Arch Phys Med Rehabil 2000,81:1370-1379.

40. O'Sullivan P, Twomey L, Allison G, Sinclair J, Miller K: Altered pat-terns of abdominal muscle activation in patients withchronic low back pain. Aust J Physiother 1997, 43:91-98.

41. Cholewicki J, Juluru K, McGill SM: Intra-abdominal pressuremechanism for stabilizing the lumbar spine. J Biomech 1999,32:13-17.

42. Sarti MA, Monfort M, Fuster MA, Villaplana LA: Muscle activity inupper and lower rectus abdominus during abdominal exer-cises. Arch Phys Med Rehabil 1996, 77:1293-1297.

43. Pope MH, Andersson GB, Broman H, Svensson M, Zetterberg C:Electromyographic studies of the lumbar trunk musculatureduring the development of axial torques. J Orthop Res 1986,4:288-297.

44. Callaghan JP, Gunning JL, McGill SM: The relationship betweenlumbar spine load and muscle activity during extensor exer-cises. Phys Ther 1998, 78:8-18.

45. Arokoski JP, Valta T, Kankaanpaa M, Airaksinen O: Activation oflumbar paraspinal and abdominal muscles during therapeu-tic exercises in chronic low back pain patients. Arch Phys MedRehabil 2004, 85:823-832.

46. Schultz A, Haderspeck K, Warwick D, Portillo D: Use of lumbartrunk muscles in isometric performance of mechanicallycomplex standing tasks. J Orthop Res 1983, 1:77-91.

47. Schultz A, Cromwell R, Warwick D, Andersson G: Lumbar trunkmuscle use in standing isometric heavy exertions. J Orthop Res1987, 5:320-329.

48. Ng JK, Parnianpour M, Richardson CA, Kippers V: Functional rolesof abdominal and back muscles during isometric axial rota-tion of the trunk. J Orthop Res 2001, 19:463-471.

49. Granata KP, Orishimo KF: Response of trunk muscle coactiva-tion to changes in spinal stability. J Biomech 2001, 34:1117-1123.

50. Janevic J, Ashton-Miller JA, Schultz AB: Large compressivepreloads decrease lumbar motion segment flexibility. JOrthop Res 1991, 9:228-236.

51. Thelen DG, Schultz AB, Ashton-Miller JA: Co-contraction of lum-bar muscles during the development of time-varying triaxialmoments. J Orthop Res 1995, 13:390-398.

52. Marras WS, Granata KP: A biomechanical assessment andmodel of axial twisting in the thoracolumbar spine. Spine1995, 20:1440-1451.

53. Macintosh JE, Bogduk N: The morphology of the lumbar erectorspinae. Spine 1987, 12:658-668.

54. McGill SM: The influence of lordosis on axial trunk torque andtrunk muscle myoelectric activity. Spine 1992, 17:1187-1193.

55. Marras WS, Davis KG, Kirking BC, Granata KP: Spine loading andtrunk kinematics during team lifting. Ergonomics 1999,42:1258-1273.

56. Van Herp G, Rowe P, Salter P, Paul JP: Three-dimensional lumbarspinal kinematics: a study of range of movement in 100healthy subjects aged 20 to 60+ years. Rheumatology 2000,39:1337-1340.

57. Vogt L, Pfeifer K, Portscher AM, Banzer W: Influences of nonspe-cific low back pain on three-dimensional lumbar spine kine-matics in locomotion. Spine 2001, 26:1910-1919.

58. Adams MA, Dolan P, Hutton WC: The lumbar spine in backwardbending. Spine 1988, 13:1019-1026.

59. el-Bohy AA, Yang KH, King AI: Experimental verification of facetload transmission by direct measurement of facet laminacontact pressure. J Biomech 1989, 22:931-941.








































































































Publish with BioMed Central and every scientist can read your work free of charge

"BioMed Central will be the most significant development for disseminating the results of biomedical research in our lifetime."

Sir Paul Nurse, Cancer Research UK

Your research papers will be:

available free of charge to the entire biomedical community

peer reviewed and published immediately upon acceptance

cited in PubMed and archived on PubMed Central

yours — you keep the copyright

Submit your manuscript here:http://www.biomedcentral.com/info/publishing_adv.asp

BioMedcentral

60. Pearcy M, Portek I, Shepherd J: The effect of low-back pain onlumbar spinal movements measured by three-dimensionalX-ray analysis. Spine 1985, 10:150-153.

61. Pearcy MJ, Tibrewal SB: Axial rotation and lateral bending in thenormal lumbar spine measured by three-dimensional radi-ography. Spine 1984, 9:582-587.

62. Hickey DS, Hukins DW: Relation between the structure of theannulus fibrosus and the function and failure of the interver-tebral disc. Spine 1980, 5:106-116.

63. Adams MA, Hutton WC: The relevance of torsion to themechanical derangement of the lumbar spine. Spine 1981,6:241-248.

64. Farfan HF, Cossette JW, Robertson GH, Wells RV, Kraus H: Theeffects of torsion on the lumbar intervertebral joints: therole of torsion in the production of disc degeneration. J BoneJoint Surg Am 1970, 52:468-497.

65. Andersson GB: The epidemiology of spinal disorders. In TheAdult Spine Edited by: Frymoyer JW. New York: Raven Press;1991:107-147.

66. Kumar S: The epidemiology and functional evaluation of low-back pain: a literature review. Eur J Phys Med Rehabil 1994,4:15-27.

67. Dalen N, Olsson KE: Bone mineral content and physical activ-ity. Acta Orthop Scand 1974, 45:170-174.

68. Nilsson BE, Westlin NE: Bone density in athletes. Clin Orthop RelatRes 1971, 77:179-182.

69. Granhed H, Jonson R, Hansson T: The loads on the lumbar spineduring extreme weight lifting. Spine 1987, 12:146-149.

70. Adams MA, Dolan P, Hutton WC: Diurnal variations in thestresses on the lumbar spine. Spine 1987, 12:130-137.

71. Roy SH, De Luca CJ, Snyder-Mackler L, Emley MS, Crenshaw RL,Lyons JP: Fatigue, recovery, and low back pain in varsity row-ers. Med Sci Sports Exerc 1990, 22:463-469.


























http://www.biomedcentral.com/info/publishing_adv.asp


sports medicine, arthroscopy, rehabilitation, therapy & … · 2017-08-29 · biomed central...

Documents